Flipping-based Policy for Chance-Constrained Markov Decision Processes

Safe reinforcement learning (RL) is a promising approach for many real-world decision-making problems where ensuring safety is a critical necessity.